Contents

RPI Camera Module

Contents

RPI Camera Module#

The Raspberry Pi Foundation provides support for the use of Raspberry Pi Camera modules on various devices, including the Raspberry Pi 4B. Documentation can be found here.

High resolution images require additional gpu_mem. This can be set using the command line raspi-config under performance options.

PiCamera#

PiCamera is a pure Python library providing access to the Raspberry Pi Camera module. PiCamera is included in the default Raspbian distribution.

# test for camera presence
import picamera

with picamera.PiCamera() as cam:
    print(cam.revision)

imx477

Camera preview#

Camera preview combined with a digital zoom features provides a useful means of acheiving critical focus. Documentation for PiCamera.zoom = (a, b, c, d), however, is sketchy. With a bit of experimentation, the parameters appear to represent

a, b: Upper left corner of the zoom frame measured as fraction of the overall frame. Should be floating point numbers between 0 and 1.
c, d: The extent of the zoom frame measured in the same units between 0 and 1.

# preview image on screen -- focus camera
import picamera
import time

with picamera.PiCamera() as camera:
    camera.start_preview()
    time.sleep(1)
    
    # zoom in effect
    for k in range(0, 950):
        camera.zoom = (k/2000, k/2000, (1000-k)/1000, (1000-k)/1000)
        time.sleep(0.01)
    time.sleep(20)

# capturing images to files
import picamera
import time

N = 5
with picamera.PiCamera() as camera:
    camera.start_preview()
    time.sleep(2)
    tic = time.time()
    for k in range(0, N):
        camera.resolution = (4056, 3040)
        camera.capture(f"test{k}.jpg")
    toc = time.time()
    print(f"fps = {(toc-tic)/N}")

fps = 1.6322479248046875

Capture to stream and display#

# capture high resolution image to a stream
from io import BytesIO
import picamera
import time
import PIL.Image

# capture image to stream
stream = BytesIO()
with picamera.PiCamera() as camera:
    camera.resolution = (4056, 3040)
    camera.start_preview()
    time.sleep(2)
    camera.capture(stream, format="jpeg")

# rewind stream
stream.seek(0)

# display image
with PIL.Image.open(stream) as im:
    display(im)

../../_images/08.03-rpi-picamera_test_7_0.png

Camera Settings#

import picamera
import time

with picamera.PiCamera() as camera:
    camera.start_preview()
    time.sleep(1)
    print(camera.resolution)
    print(camera.exposure_mode)
    print(camera.meter_mode)
    print("Shutter speed = ", camera.exposure_speed/1000000, "secs")
    print(camera.analog_gain)
    print(camera.digital_gain)
    print(camera.iso)
    print(camera.awb_gains)

800x480
auto
average
Shutter speed =  0.032979 secs
8
365/256
0
(Fraction(243, 128), Fraction(369, 128))

import picamera
import time
import PIL.Image

with picamera.PiCamera() as camera:
    camera.start_preview()
    camera.resolution = "1080p"
    camera.shutter_speed = 200000
    camera.iso = 800
    camera.capture("mypic.png")
    
with PIL.Image.open("mypic.png") as im:
    display(im)

../../_images/08.03-rpi-picamera_test_10_0.png

Capture to numpy array#

import time
import picamera
import numpy as np

with picamera.PiCamera() as camera:
    camera.resolution = (1920, 1088)
    camera.framerate = 24
    output = np.empty((1088, 1920, 3), dtype=np.uint8)
    camera.capture(output, 'rgb')

import matplotlib.pyplot as plt
fig, ax = plt.subplots(1, 1, figsize=(12, 8))
ax.imshow(output)

<matplotlib.image.AxesImage at 0xa14cfd10>

../../_images/08.03-rpi-picamera_test_13_1.png

import time
import picamera
import numpy as np
import cv2
import matplotlib.pyplot as plt

with picamera.PiCamera() as camera:
    camera.resolution = (1920, 1088)
    camera.framerate = 24
    img = np.empty((1088, 1920, 3), dtype=np.uint8)
    camera.capture(img, 'bgr')
    img = img.reshape((1088, 1920, 3))
    

img_gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
ret, thresh = cv2.threshold(img_gray, 150, 255, cv2.THRESH_BINARY)
plt.imshow(thresh)

contours, hierarchy = cv2.findContours(image=thresh, mode=cv2.RETR_TREE, method=cv2.CHAIN_APPROX_NONE)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/tmp/ipykernel_982/1894840504.py in <module>
      3 plt.imshow(thresh)
      4 
----> 5 contours, hierarchy = cv2.findContours(image=thresh, mode=cv2.RETR_TREE, method=cv2.CHAIN_APPROX_NONE)

ValueError: too many values to unpack (expected 2)

../../_images/08.03-rpi-picamera_test_15_1.png

img_blur = cv2.GaussianBlur(img_gray, (3,3), 0)
edges = cv2.Canny(image=img_blur, threshold1=100, threshold2=300)
fig, ax = plt.subplots(1, 1, figsize=(12, 8))
ax.imshow(edges)

<matplotlib.image.AxesImage at 0xa1283890>

../../_images/08.03-rpi-picamera_test_16_1.png

#

contours, hierarchy = cv2.findContours(image=thresh, mode=cv2.RETR_TREE, method=cv2.CHAIN_APPROX_NONE)
image_copy = image.copy()
6
cv2.drawContours(image=image_copy, contours=contours, contourIdx=-1, color=(0, 255, 0), thickness=2, lineType=cv2.LINE_AA)
7
                
8
# see the results
9
cv2.imshow('None approximation', image_copy)
10
cv2.waitKey(0)
11
cv2.imwrite('contours_none_image1.jpg', image_copy)
12
cv2.destroyAllWindows()